Update 1: by a skeptical comment I RAW comparison shots of my first copy again individually in Lightroom optimized (CA): the center of the picture is so sharp and contrasty that a tearful eyes. Worlds better than 28 and 18-200. That had not been my problem: already in the semi-field (a little over half way from the center in a corner), the picture varies greatly: the Chrom.Aberrationen can be manually actually largely correct (the automatic in 3.4 LR always enough not yet - on the basis of my first test), but the focus can then be strong after - and well below that of said comparing lenses.
Because the center I change to 2 star - but more is not in it, because the total power over the area in my opinion is well below that of the 18-200VR. I will soon get other copies - then there's another update. Possibly I had a bad piece.
Update 2: I've now tested a second copy. That was better. The CA's edge were not still completely compensated, but no comparison with my first piece. For normal purposes totally ok. This specimen of 1.8 / 35 leaves unlike my first the 18-200VRII far behind.
I have to the 2.0 / 35 tested in parallel. This is a full frame lens, but for DX also suitable. The 2.0 / 35 is even better for my needs - because it brings the more consistent performance over the entire area.
So I change my vote for the 1.8 / 35 on 3 stars. I have but ultimately the 2.0 / 35 reserve because I am concerned with landscapes in which performance over the entire image area is important.
[...]